The Study of Multi-source Image Fusion
碩士 === 國防大學中正理工學院 === 電子工程研究所 === 94 === Recently, image fusion technique has been become to an important image analysis and computer vision technique, there are many applications of image fusion, some of which include medical imaging, remote sensing, nighttime target identify, multi-focus imaging,...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Language: | zh-TW |
Published: |
2006
|
Online Access: | http://ndltd.ncl.edu.tw/handle/04198525475602249941 |
id |
ndltd-TW-094CCIT0428037 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-TW-094CCIT04280372015-10-13T10:34:48Z http://ndltd.ncl.edu.tw/handle/04198525475602249941 The Study of Multi-source Image Fusion 多重來源影像融合之研究 Kao,Li-Tsun 高儷純 碩士 國防大學中正理工學院 電子工程研究所 94 Recently, image fusion technique has been become to an important image analysis and computer vision technique, there are many applications of image fusion, some of which include medical imaging, remote sensing, nighttime target identify, multi-focus imaging, and military affairs application, which has the widely applied foreground. A general definition of image fusion has been stated as “the combination of two or more different images to form a new image by using a certain algorithm”. Since how to integrate these distinct multi-source images provide different image data becomes a valuable research topic. The image fusion with multi-resolution analysis method has been developed to a usual approach. According to the exploiting the theory of wavelet, the researches toward to fusion the image with the orthogonal wavelet transform (OWT). In addition, also some researches utilized the discrete cosine transform (DCT) domain to define the contrast and perform the fusion. However, these methods face to the distortion problem, that is, the appearance of the artifact. This thesis provides a novel image fusion for the multi-focus images. Firstly, we apply the smoothing filter intensity modulation (SFIM) to enhance the detail of input images. Then, by the exploit of the theory of wavelet, we develop the definition of the contrast in non-orthogonal wavelet domain to perform the selection of feature. Finally, a fusion image is gained that suit to human visual system more than the original image. Experimental results demonstrate that the proposed approach can accomplish a more detailed representation of the obtained scene. Furthermore, this approach can achieve better fusion performance than the fusion approaches of the standard OWT and DCT. Tu,Te-Ming Chang,Chien-Ping 杜德銘 張劍平 2006 學位論文 ; thesis 46 zh-TW |
collection |
NDLTD |
language |
zh-TW |
format |
Others
|
sources |
NDLTD |
description |
碩士 === 國防大學中正理工學院 === 電子工程研究所 === 94 === Recently, image fusion technique has been become to an important image analysis and computer vision technique, there are many applications of image fusion, some of which include medical imaging, remote sensing, nighttime target identify, multi-focus imaging, and military affairs application, which has the widely applied foreground. A general definition of image fusion has been stated as “the combination of two or more different images to form a new image by using a certain algorithm”. Since how to integrate these distinct multi-source images provide different image data becomes a valuable research topic.
The image fusion with multi-resolution analysis method has been developed to a usual approach. According to the exploiting the theory of wavelet, the researches toward to fusion the image with the orthogonal wavelet transform (OWT). In addition, also some researches utilized the discrete cosine transform (DCT) domain to define the contrast and perform the fusion. However, these methods face to the distortion problem, that is, the appearance of the artifact. This thesis provides a novel image fusion for the multi-focus images. Firstly, we apply the smoothing filter intensity modulation (SFIM) to enhance the detail of input images. Then, by the exploit of the theory of wavelet, we develop the definition of the contrast in non-orthogonal wavelet domain to perform the selection of feature. Finally, a fusion image is gained that suit to human visual system more than the original image.
Experimental results demonstrate that the proposed approach can accomplish a more detailed representation of the obtained scene. Furthermore, this approach can achieve better fusion performance than the fusion approaches of the standard OWT and DCT.
|
author2 |
Tu,Te-Ming |
author_facet |
Tu,Te-Ming Kao,Li-Tsun 高儷純 |
author |
Kao,Li-Tsun 高儷純 |
spellingShingle |
Kao,Li-Tsun 高儷純 The Study of Multi-source Image Fusion |
author_sort |
Kao,Li-Tsun |
title |
The Study of Multi-source Image Fusion |
title_short |
The Study of Multi-source Image Fusion |
title_full |
The Study of Multi-source Image Fusion |
title_fullStr |
The Study of Multi-source Image Fusion |
title_full_unstemmed |
The Study of Multi-source Image Fusion |
title_sort |
study of multi-source image fusion |
publishDate |
2006 |
url |
http://ndltd.ncl.edu.tw/handle/04198525475602249941 |
work_keys_str_mv |
AT kaolitsun thestudyofmultisourceimagefusion AT gāolìchún thestudyofmultisourceimagefusion AT kaolitsun duōzhòngláiyuányǐngxiàngrónghézhīyánjiū AT gāolìchún duōzhòngláiyuányǐngxiàngrónghézhīyánjiū AT kaolitsun studyofmultisourceimagefusion AT gāolìchún studyofmultisourceimagefusion |
_version_ |
1716830569401155584 |